产前保健服务利用指标的设计效应初探

Design effects for the survey on antenatal care utilization in 12 rural counties of China

目的计算整群抽样调查所得产前保健服务利用指标的设计效应，为今后妇幼卫生领域的相关调查设计提供参考。方法利用联合国儿童基金会“母子健康综合项目”2011年9月-10月对中国西部6省／自治区12县2岁以下儿童母亲调查得到的数据，根据调查的抽样方法，计算各产前保健服务利用指标的设计效应。结果被调查地区的各产前保健服务利用指标的设计效应在2．8—4．9之间。反映产前检查内容的内容合格率和至少接受4项检查比例两个指标的设计效应（分别为4．64、4．90）大于其他指标。结论整群抽样时指标的设计效应较大，不同产前保健服务利用指标、不同县同一指标的设计效应不同。在设计整群抽样调查时，需要根据设计效应，调整样本量，以保证样本统计量的精确度。

Objective The purposes of this study were to estimate the design effects for the indicators assessing antenatal care utilizations from a duster survey, and to provide the design effects as reference for calculating sample size in future surveys. Methods Data were derived from a household survey to mothers of children under two in 12 rural coun- ties of China during September to October in 2011. According to the cluster sampling design of survey, design effects of in- dicators for antenatal care utilization were calculated. Results Design effects for antenatal care utilization indicators were in the range of 2.8-4.9. The content of antenatal care and the percentage pregnancy received at least 4 items, with the two indicators measuring the quality of antenatal care content having had larger design effects （4.64 and 4.90, respectively） than others. Conclusion The cluster samples likely had larger design effects. Design effects varied substantially among different counties, as well as by different indicators of antenatal care utilization. These findings suggest that the cluster sur- veys might substantially increase sample size according to design effect to maintain adequate statistical power.